Autonomous Robotics from the Ground Up

  • BrandElektor Publishing

Description

The Black Shadow Project: From Sensors to Embedded AI

Most robotics books stop at individual components or jump to frameworks that hide the hardware. This book explores how a complete autonomous robot is designed, integrated and debugged from first principles, following an engineer’s practical problems, failed approaches and workable solutions.

Three dedicated dual-core microcontrollers divide perception, motor control and navigation, reducing timing conflicts. Sixteen time-of-flight sensors across eight independent I²C buses provide 360° environmental coverage at 20 Hz. Their measurements feed a trained neural network that performs real-time obstacle avoidance directly on the microcontroller using TensorFlow Lite for Microcontrollers, without cloud services or a GPU.

A built-in Wi-Fi access point and browser interface give your smartphone access to live telemetry, sensor visualisation and manual driving without installing an app.

An onboard virtual machine executes bytecode scripts one instruction per scheduler tick, allowing multi-step manoeuvres without blocking real-time control loops. This establishes a foundation for increasingly independent behaviour.

The closing chapter examines what separates shared control from the book’s goal of Level 3 autonomy: self-localisation, mission planning, reliable heading, predictive power management and recognising when to stop. It outlines possible solutions, then challenges you to design Black Shadow 2.0. Throughout, the emphasis remains on integration, troubleshooting and practical engineering decisions.


Product form

The Black Shadow Project: From Sensors to Embedded AI Most robotics books stop at individual components or jump to frameworks that... Read more

€ 39,95 (incl. VAT)
Members: € 35,96

    Details

    SKU: 21360
    EAN: 9783895767807
    ISBN: 978-3-89576-780-7
    Author: Jan De Sutter
    Language: English
    Pages: 326 (Full Color)
    Format: 17 x 24 cm (Paperback)

    Description

    The Black Shadow Project: From Sensors to Embedded AI

    Most robotics books stop at individual components or jump to frameworks that hide the hardware. This book explores how a complete autonomous robot is designed, integrated and debugged from first principles, following an engineer’s practical problems, failed approaches and workable solutions.

    Three dedicated dual-core microcontrollers divide perception, motor control and navigation, reducing timing conflicts. Sixteen time-of-flight sensors across eight independent I²C buses provide 360° environmental coverage at 20 Hz. Their measurements feed a trained neural network that performs real-time obstacle avoidance directly on the microcontroller using TensorFlow Lite for Microcontrollers, without cloud services or a GPU.

    A built-in Wi-Fi access point and browser interface give your smartphone access to live telemetry, sensor visualisation and manual driving without installing an app.

    An onboard virtual machine executes bytecode scripts one instruction per scheduler tick, allowing multi-step manoeuvres without blocking real-time control loops. This establishes a foundation for increasingly independent behaviour.

    The closing chapter examines what separates shared control from the book’s goal of Level 3 autonomy: self-localisation, mission planning, reliable heading, predictive power management and recognising when to stop. It outlines possible solutions, then challenges you to design Black Shadow 2.0. Throughout, the emphasis remains on integration, troubleshooting and practical engineering decisions.


    Login

    Forgot password?

    Don't have an account yet?
    Create account